Standard Alignment Tapes – A History At and After Ampex1

Jay McKnight
Magnetic Reference Laboratory
San Jose, California
jay.mck@comcast.net

0 INTRODUCTION
Ampex announced the Model 200A tape recorder in 1947 October. This was the first USA-made
professional audio tape recorder with the quality and durability needed for broadcasting and
sound-recording studio operation. It was first used for recording, editing, and playing the Bing Crosby
radio show, and also for time-zone delaying radio programs broadcast by the networks across the USA.
For such professional applications, Ampex design engineers (who were already familiar with
disc-recording standards) understood that the recorders and reproducers, and the recordings made on
different machines, needed to not only be of the highest quality but also interchangeable to ensure
consistent equalization, azimuth, and recorded level. The Model 200A manual [1] notes (p 3) “Using the
best possible input and monitoring equipment, recordings cannot be distinguished from original program
material when directly compared.”
By the late 1940s, the use of standards for recording and reproducing phonograph discs had been in use
for twenty years. The earliest sound recorder that we would consider to be “high fidelity” was the
electromechanical phonograph disc-recording system developed by the Bell Telephone Laboratories in the
1920s [2] . Those engineers realized that one requirement for a professional system was interchangeability
of the recorders, the phonograph discs, and the reproducers. To that end, they defined a quantity for the
recorded signal, called the recorded groove velocity, and developed means for measuring it, which later
became an IRE (IEEE) Standard. Once these standards were defined, Bell engineers developed test
recordings that contained the standard levels and the standard frequency responses. To standardize a
recording and reproducing system, you would simply play the test record and adjust your reproducer for
standard output voltage and flat frequency response. Only after you did that would you adjust your
recording system to ensure that the standard reproducer again provided the standard output voltage and flat
frequency response. Assuming no further changes were made, you could then record an audio program with
assurance that it would conform to the standard.
The situation with a tape recorder is exactly the same – and on page 10 of the Ampex 200A manual
we find: “each machine is checked to match a standard tape to insure that recordings will be
interchangeable among machines.”

1 THE FIRST ALIGNMENT TAPES AT AMPEX

To make the first alignment tape, Ampex engineers calibrated a specific recorder to record at optimum
frequency response and level. Using this recorder, they then made a single “Standard Alignment Tape”
with various tones at standard levels—this tape was then used to adjust playback level and equalization on
all tape recorders ready for distribution. But having only one standard alignment tape is a bad idea since
tapes are easily lost, erased, or otherwise damaged.

1
Originally presented at the Association for Recorded Sound Collections (ARSC) Annual Conference at
Stanford University, 2008 Mar 26...29

Page 1
 The Model 200A manual (page 19) mentions a “standard tape” with 18 tones across the audio band for
setting up the azimuth, gain, and frequency response of the reproducer, with “taps” used to identify the
frequencies. The exact history is unclear, but we do know that Ampex engineer Frank Lennert recorded a
“master” standard tape and set up a pair of Model 200A recorders for copying. We also know that Alex
Poniatoff hired “a bright high-schooler” (Ray Dolby, later famous for his noise-reduction system) to make
duplicate tapes from this original master tape. Ray’s memory was that this was in 1949 June. It is unclear
whether these tapes were given or sold to Model 200A owners; however, the Model 200A manual (dated
1948 July 15) describes adjusting the reproducer “from a standard tape” and assumes its availability.
The recording level and the equalization standards used on the Model 200A were designed for use with
3M Type 112 magnetic tape. This tape had a randomly-oriented coating [3]. An improved tape, 3M Type
111 (yes, the numbering sequence is reversed) was used, beginning with the Ampex Model 300 recorder.
3M Type 111, like all subsequent magnetic tapes designed for audio recording (by both 3M and other
manufacturers), had a longitudinally-oriented coating; this provided 3 dB higher output than 3M Type 112
[4] .
With the switch to the new improved 3M Type 111 tape, the Ampex model 200A level and equalization
standards were obsolete once the model 300 was developed. That the original standard was abandoned isn’t
historically all that significant because few original tapes recorded on an Ampex 200A survive. To
understand why this is so, one must remember that introduction of high-quality magnetic recording in 1948
generated much excitement—not only because of its wide bandwidth and low noise but also because it was
so much more cost-effective than instantaneous cellulose-nitrate disc recording. As a result, most tapes
recorded on Ampex 200A machines were reused after the recorded programing was broadcast.
The Model 200A equalization and level were rather arbitrarily optimized. This presented two
disadvantages. First, there is no way to write a specification and standard for your recorded tapes or for the
recorder. Second, if your “Standard Tape” becomes erased there is no way to go back and reconstruct it (to
do that, you would need a fundamental physical specification).
By 1948 May, Ampex was developing the Model 300. The first Model 300 Manual has pages dated
1949 August 15, as does the Manual for the improved Model 200A (called the Model 201). The 300
Manual mentions the standards proposed by the NAB Subcommittee on Magnetic Recording, so that
standards committee must have existed before mid-1949.
Frank Lennert developed a simplified equalization for Ampex that later became the NAB
equalizations for 7.5 and 15 in/s (which are still used in the USA). In the Manual for the Ampex 201,
Lennert published a comparison between the equalizations used on several professional recorders at 15
in/s—see page 11 of the Ampex201 manual [5] . It shows that by that time, all of the 15 in/s machines
(except the Magnecord) used essentially the same equalization.
By 1950, magnetic tape recording was widely used around the world. The International Radio
Consultative Committee (known by its French initials CCIR) was studying the means for standardizing the
frequency response of the reproducers by fundamental methods, with participation by several European
countries. They published a number of reports that were summarized in a 1952 AES convention paper by
John D. Bick of RCA [6] .
Over the years, there has been confusion about the name of the physical quantity that describes the signal
recorded on magnetic tape. In the original CCIR reports, surface induction was used (“induction” is an
alternate name for “magnetic flux density”) because this quantity (like the output from a magnetic
reproducing head) has an output voltage versus frequency that rises 6 dB per octave from a constant flux
input.
John D. Bick titled his 1953 paper Methods of Measuring Surface Induction of Magnetic Tape but he
soon found out that the recorded quantity in his paper was actually flux rather than flux density, and in fact
published a little-known Errata [7] in the AES Journal in 1954 April, acknowledging that this error—that is,
the use of the term “surface induction” instead of “flux”—had been incorporated into CCIR and NAB

Page 2
standards and “will be corrected in the near future.” Unfortunately it was not corrected for many years; we
still get occasional inquires “why does MRL say flux? Isn’t it flux density?” And the answer is, no, it really
is just plain “flux.”
Lennert at Ampex gave a technical paper at an IRE Convention in 1952 and published it in an IRE
Transactions in 1952 [8] . It explained the equalizations and the basis for them, and the notion of measuring
the tape flux using an “ideal” reproducing head. This measurement method was used by Ampex to calibrate
the reproducers that were used to make the “Ampex Standard Alignment Tapes,” and, in essence, is still
used to this day.
The first Ampex specification of the “operating level” (which we now call “reference fluxivity”) was
the flux level that would produce one percent harmonic distortion on playback, which is approximately 6
dB below the level for three percent harmonic distortion. This is not a very repeatable flux because it
depends on the particular tape used for the measurement and on the way that the biasing current is set.
Furthermore, the “fundamental cancelling” distortion meters used at the time measured the tape noise and
bias leakage along with the distortion products; this could cause an error in the measured distortion. Thus to
avoid additional measurement error, the thinking at the time was that it was better to copy a standard flux
from a reference recording rather than being re-measured “from scratch” and a new standard tape recorded
each time it was needed. The Rundfunktechniche Institut of the German Broadcasting System developed a
flux-measuring standard in 1956 [9] . We revisited this standard and improved it in AES papers in 1970 [10]
and in 1998 [11]; later, we wrote an AES Flux Measurement Standard in 1982 and updated it in 2000 [12]
The first Ampex standard alignment tapes, as I said before, were just copies of a master tape. This has
several serious drawbacks. First, whatever level dropouts are on the master are copied, making the levels
on the copy less stable. Second, any frequency-response errors on the master are copied. And third, a
master tape that is repeatedly handled and played may suffer partial erasure, introducing an error in the
master that also gets copied. It is therefore much better to record the tones on the Standard Alignment Tape
directly from an oscillator and re-record only the voice announcements from another tape.

2 THE AMPEX “STANDARD TAPE LAB”

Around 1949, Ampex developed an improved system for making a Standard Tape. It used a master
voice-announcement tape (played from another recorder) and combined these announcements with tones
from a multi-frequency oscillator. An operator switched the record input to the machine recording the
newly-created master tape between voice announcements and tones. The record level was trimmed
manually, as needed.
This system is shown in a 1958 paper by Frank Richards in the AES Journal [13]. It includes photos of
the modified Model 300 used at Ampex to record the Standard Tapes. Also shown is the playback head
azimuth indicator on this transport, the specially-built oscillator (with manually switchable frequencies), a
switch to change between voice and tone, and a level trim-pot for each frequency. This system was used at
Ampex from 1950 until the 1980s. A copy of this system was later used at both STL (Standard Tape
Laboratory) and MRL (Magnetic Reference Laboratory).
Richardson also describes the manufacturing process:
The voice instructions of the Standard Tape are reproduced from another machine, and
since all Standard Tapes are recorded in reverse to eliminate the necessity for rewinding,
the voice is reproduced backwards from one track of a two-channel machine. The other
track carries cue tones to guide the operator who, incidentally, is a laboratory engineer
devoting full time to the production of these tapes. Surveillance of each tape is maintained
as it is being recorded, and any tape showing deviation from accepted standards is
immediately discarded. Tolerances of ± 2/3 minutes for azimuth setting, and ± 1/4 dB in
level are maintained.

Page 3
 Let me add a comment here: We at MRL think that recording the Standard Tapes backwards, as Ampex
did, was actually not such a good idea. Here’s why:
 Any level corrections from one tone to the next occur at the end of the tone. But when you’re
adjusting the response of a reproducer, you are probably making the last fine adjustment at the
end, when the recorded level is moving around. Better to record in the “normal play” direction,
so the corrections are made at the beginning of the tone so that the end of the tone is steady.
 When the recording is “heads out” you have to rewind it after use for storage. Rewinding
(typically a very fast, high-tension rewind) is not as good for storage as a slower playback wind.
It’s better to store the tape played then only rewound just before use; this way, the master tape
will be loosened up with a rewind but only just before use.
 Even though engineers who are used to “heads out” tapes may be confused by a calibration tape
that is stored “tails out,” the benefits of storing an expensive calibration tape this way far
outweigh any perceived inconvenience.
Continuing research at Ampex and other companies improved the understanding and practices for the
standardization of the tape flux (“recorded level”), the flux vs frequency (“equalization”), and the azimuth
[14] . These also resulted in papers in the AES Journal about using Standard Alignment Tapes by Morrison
[15] and by McKnight [16] .

3 COMPETITION
Morrison [17] left Ampex to found the competing “Standard Tape Laboratory” in 1968; STL closed
in1990. McKnight and Bardakos left Ampex to found “Magnetic Reference Laboratory” in 1972; we are
still making Calibration Tapes in 2019. The MRL system that we developed in the 1980s is computer
automated and now has the voice on a hard-disk recording, with computer-controlled switching of voice
and tone, and an automatic gain control to keep a constant recorded flux even if the tape sensitivity varies
a little during the recording.
The Ampex Standard Tapes had one format—announcements and standardized tones and levels. At
MRL, we have experienced considerable demand for tapes with more (or fewer tones), shorter or longer
duration, and customized features including sweeps and pink or white noise. With the computer control,
we have the ability to program many combinations of voice-announced tones.
Standards for sound recording are now written and published by the International Electrotechnical
Commission (IEC). The standard for magnetic tape recording (originally issued as IEC Publication 94) is
now called IEC Publication 60 094 and is in several parts. Part 2 is “Magnetic tape recording and
reproducing systems, Calibration Tapes” which gives specifications for standard Calibration Tapes
(including frequencies, levels, and tolerances). We at MRL have adopted the IEC name “Calibration
Tapes” although many people still call them “standard alignment tapes.”
There are unfortunately no published standards for the measurement of Calibration Tapes.

4 THE VOICES
People often ask us “Who are the voices on the Standard Alignment Tapes?” I don’t have exact
dates, but here are the names and approximate times:
Around 1948, according to the Model 200A manual, there were “taps” (not voices) to identify the tone
frequencies.
Around 1948 or 1949, Frank Lennert [18] , Ampex engineer and Manufacturing Manager, is reported
to have recorded the voice for the tapes. I’ve never seen (nor heard) one of these tapes. This may be the
tape referenced in the 1949 June 15 manual.

Page 4
 The 1952 manual for the Model 300 mentions the #4494 Standard Alignment tape (quarter-inch, 15
in/s). This was packaged in a small round gray-painted metal can; the announcements were the voice of
Ross Snyder [19] , Ampex engineer and Product Manager. The 1952 December Ampex price list has this
listed for US$11.
Around 1960, a new voice track was recorded for Ampex by audio engineer Bob Morrison. The
package was also changed from the metal can to a rectangular cardboard box. The part number was
changed to 01-31321-01 for the quarter-inch, 7.5 in/s tape. I have a sample dated 1961 April.
When Morrison left Ampex to found Standard Tape Lab around 1968, the Ampex voice tape was again
newly-recorded, this time by Tony Bardakos (who was then working at the Ampex Standard Tape Lab).
Bardakos was to go on to found Magnetic Reference Lab in 1972; he then recorded all MRL voice tapes
used from 1972 to 1990 (when we introduced a new computer-controlled signal-generation system).
The current (1990…2019) MRL voice tape was recorded in 1990 at the Menlo Park (California)
studios of Music Annex by Tim Enos, a professional voiceover artist. These announcements were
formatted so that they could be flexibly assembled into the necessary announcements “on the fly” as the
Calibration Tapes are recorded.
Thanks to David Dintenfass for editing this paper.

5 REFERENCES
[1] Ampex Model 200A Manual, online at
http://www.aes.org/aeshc/docs/ampex200a/ampex-200a-instruction-book.pdf
[2] J. P. Maxwell and H. C. Harrison, “Methods of High-Quality Recording and Reproducing of Music and
Speech Based on Telephone Research,” JAES Vol 26 Nr 5, pp 327...342 (1978 May), originally published
in Trans AIEE in 1926. It is curious that the learned Bell System physicists and engineers, in inventing and
using electro-mechanical and electro-acoustical systems, never realized or admitted the convenience of
Firestone’s “mobility analog” for mechanical systems—see the Wikipedia article “Mobility analog.”
[3] R. A. von Behren, “Some Design Criteria for Magnetic Tape,” JAES Vol 3 Nr 4, pp. 210...216
(1955 Oct)
[4] 3M Tapes, technical data at http://www.aes.org/aeshc/docs/3mtape/aorprod-si.pdf
[5] Ampex Model 201 Manual, online at
http://www.aes.org/aeshc/docs/ampex200a/ampex-201-conversion.pdf
[6] J. D. Bick, “Methods of Measuring Surface Induction of Magnetic Tape,” JAES Vol 1 Nr 1, pp. 4...9
(1953 Jan); (actually surface flux as published in a correction at JAES, originally at [7], now included with
the online reprint of the original paper)
[7] J. D. Bick, Errata to “Methods of Measuring Surface Induction of Magnetic Tape,” JAES Vol 2 Nr 2
(1954 Apr), p. 112, 113; now incorporated into [6]
[8] F. G. Lennert, “Equalization of Magnetic Tape Recorders for Audio and Instrumentation
Applications,” Trans IRE- PGA, Vol AU-1, Nr 2 (1953 Mar-Apr); online at
http://mrltapes.com/lennert_eq-of-mag-tape.pdf
[9] Otto Schmidbauer, “Determining the Magnetization of Magnetic Tape,” JAES Vol 46 Nr 10 (1998 Oct),
pp. 859…864; online at http://mrltapes.com/schmidbauer_determining-magnetization.pdf
[10] J. G. McKnight, “Tape Flux Measurement Theory and Verification,” JAES Vol 18 Nr 3 pp. 250...262
(1970 May); online at http://mrltapes.com/mcknight_tape-flux-measurement-theory.pdf

Page 5
[11] John G. (Jay) McKnight, Benito E. Cortez, Jeffrey A. McKnight, “Tape Flux Measurement
Revisited,” JAES Vol 46 Nr 10 pp. 845...858 (1998 Sept); online at
http://mrltapes.com/mcknight_tape-flux-measurement-revisited.pdf
[12] AES7-2000 (s2012): AES standard for the preservation and restoration of audio recording—Method
of measuring recorded fluxivity of magnetic sound records at medium wavelengths (Revision of
AES7-1982)
[13] Frank F. Richards,“The Standard Tape as a Professional Audio Tool,” JAES Vol 6 Nr 3, pp.
193...196, 208 (1958 Jun)
[14] John G. McKnight, “Flux and Flux-frequency Measurements and Standardization in
Magnetic Recording,” JSMPTE, Vol. 78, pp. 457...472 (1969 Jun); online at
http://mrltapes.com/mcknight_flux-and-flux-frequency-response-measurements.pdf
[15] Robert K. Morrison, “Reproducer Test Tapes, Their Evolution and Manufacture,” JAES Vol 15
Nr 2, pp. 157...162 (1967 Mar) ; online at http://mrltapes.com/morrison_reproducer-test-tapes.pdf
[16] John G. McKnight, “Tape Reproducer Response Measurements with a Reproducer Test Tape,”
JAES Vol 15 Nr 2, pp. 152...156 (1967 Mar) ; online at
http://mrltapes.com/mcknight_tape-reproducer-response.pdf
[13] Robert Keith Morrison obituary
[18] Frank Lennert obituary
[19] Ross Snyder obituary

Page 6